In [40]:
import numpy as np
import scipy as sp
import matplotlib as mpl
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
%matplotlib inline

import tensorflow as tf
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.applications import VGG16
from tensorflow.keras import layers
from tensorflow.keras import models
from tensorflow.keras import regularizers,optimizers
from keras.models import Sequential
from tensorflow.keras.layers import Flatten
from tensorflow.keras.layers import Dropout
from tensorflow.keras.layers import Dense
from tensorflow.keras.layers import Input
from tensorflow.keras.models import Model
from keras.layers import Conv2D, MaxPooling2D,Activation
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.utils import to_categorical
from tensorflow.keras.preprocessing.image import img_to_array
from tensorflow.keras.preprocessing.image import load_img
from tensorflow.keras.preprocessing.image import array_to_img
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelBinarizer
import pickle              # For saving models to disk and loading saved models
import os
import zipfile
import PIL
import PIL.Image
import pathlib
import cv2 as cv
In [41]:
# Initializing the random number generator
import random
random.seed(0)

# Ignoring the warnings
import warnings
warnings.filterwarnings("ignore")
In [42]:
path = 'C:/Users/uprak/capstone'
os.chdir(path)
In [43]:
ls

 Volume in drive C is Windows
 Volume Serial Number is D87D-3DAA

 Directory of C:\Users\uprak\capstone

12-12-2021  09:12    <DIR>          .
12-12-2021  09:12    <DIR>          ..
07-12-2021  17:05    <DIR>          .ipynb_checkpoints
07-12-2021  15:56    <DIR>          Annotations
12-12-2021  09:12         4,263,595 capstone_cv2.ipynb
09-12-2021  19:30    <DIR>          Car Images
07-12-2021  15:13             5,454 Car names and make.csv
07-12-2021  16:01         1,666,501 test.csv
07-12-2021  22:35         6,514,171 test1.csv
07-12-2021  16:01         1,696,197 train.csv
07-12-2021  22:17         7,129,237 train1.csv
               6 File(s)     21,275,155 bytes
               5 Dir(s)  98,481,082,368 bytes free
In [44]:
path1 = os.getcwd()
path1
Out[44]:
'C:\\Users\\uprak\\capstone'
In [45]:
car_img_path = os.path.join(path1,'Car Images')
car_ann_path = os.path.join(path1,'Annotations')
In [46]:
ds_car_name = pd.read_csv("Car names and make.csv",names=['Car Name'])
ds_car_name
Out[46]:
Car Name
0 AM General Hummer SUV 2000
1 Acura RL Sedan 2012
2 Acura TL Sedan 2012
3 Acura TL Type-S 2008
4 Acura TSX Sedan 2012
... ...
191 Volkswagen Beetle Hatchback 2012
192 Volvo C30 Hatchback 2012
193 Volvo 240 Sedan 1993
194 Volvo XC90 SUV 2007
195 smart fortwo Convertible 2012

196 rows × 1 columns

In [47]:
count = 0
car_dict = {}
for car_name in ds_car_name['Car Name'].values:
    car_dict[count+1] = car_name
    count=count+1
car_dict
Out[47]:
{1: 'AM General Hummer SUV 2000',
 2: 'Acura RL Sedan 2012',
 3: 'Acura TL Sedan 2012',
 4: 'Acura TL Type-S 2008',
 5: 'Acura TSX Sedan 2012',
 6: 'Acura Integra Type R 2001',
 7: 'Acura ZDX Hatchback 2012',
 8: 'Aston Martin V8 Vantage Convertible 2012',
 9: 'Aston Martin V8 Vantage Coupe 2012',
 10: 'Aston Martin Virage Convertible 2012',
 11: 'Aston Martin Virage Coupe 2012',
 12: 'Audi RS 4 Convertible 2008',
 13: 'Audi A5 Coupe 2012',
 14: 'Audi TTS Coupe 2012',
 15: 'Audi R8 Coupe 2012',
 16: 'Audi V8 Sedan 1994',
 17: 'Audi 100 Sedan 1994',
 18: 'Audi 100 Wagon 1994',
 19: 'Audi TT Hatchback 2011',
 20: 'Audi S6 Sedan 2011',
 21: 'Audi S5 Convertible 2012',
 22: 'Audi S5 Coupe 2012',
 23: 'Audi S4 Sedan 2012',
 24: 'Audi S4 Sedan 2007',
 25: 'Audi TT RS Coupe 2012',
 26: 'BMW ActiveHybrid 5 Sedan 2012',
 27: 'BMW 1 Series Convertible 2012',
 28: 'BMW 1 Series Coupe 2012',
 29: 'BMW 3 Series Sedan 2012',
 30: 'BMW 3 Series Wagon 2012',
 31: 'BMW 6 Series Convertible 2007',
 32: 'BMW X5 SUV 2007',
 33: 'BMW X6 SUV 2012',
 34: 'BMW M3 Coupe 2012',
 35: 'BMW M5 Sedan 2010',
 36: 'BMW M6 Convertible 2010',
 37: 'BMW X3 SUV 2012',
 38: 'BMW Z4 Convertible 2012',
 39: 'Bentley Continental Supersports Conv. Convertible 2012',
 40: 'Bentley Arnage Sedan 2009',
 41: 'Bentley Mulsanne Sedan 2011',
 42: 'Bentley Continental GT Coupe 2012',
 43: 'Bentley Continental GT Coupe 2007',
 44: 'Bentley Continental Flying Spur Sedan 2007',
 45: 'Bugatti Veyron 16.4 Convertible 2009',
 46: 'Bugatti Veyron 16.4 Coupe 2009',
 47: 'Buick Regal GS 2012',
 48: 'Buick Rainier SUV 2007',
 49: 'Buick Verano Sedan 2012',
 50: 'Buick Enclave SUV 2012',
 51: 'Cadillac CTS-V Sedan 2012',
 52: 'Cadillac SRX SUV 2012',
 53: 'Cadillac Escalade EXT Crew Cab 2007',
 54: 'Chevrolet Silverado 1500 Hybrid Crew Cab 2012',
 55: 'Chevrolet Corvette Convertible 2012',
 56: 'Chevrolet Corvette ZR1 2012',
 57: 'Chevrolet Corvette Ron Fellows Edition Z06 2007',
 58: 'Chevrolet Traverse SUV 2012',
 59: 'Chevrolet Camaro Convertible 2012',
 60: 'Chevrolet HHR SS 2010',
 61: 'Chevrolet Impala Sedan 2007',
 62: 'Chevrolet Tahoe Hybrid SUV 2012',
 63: 'Chevrolet Sonic Sedan 2012',
 64: 'Chevrolet Express Cargo Van 2007',
 65: 'Chevrolet Avalanche Crew Cab 2012',
 66: 'Chevrolet Cobalt SS 2010',
 67: 'Chevrolet Malibu Hybrid Sedan 2010',
 68: 'Chevrolet TrailBlazer SS 2009',
 69: 'Chevrolet Silverado 2500HD Regular Cab 2012',
 70: 'Chevrolet Silverado 1500 Classic Extended Cab 2007',
 71: 'Chevrolet Express Van 2007',
 72: 'Chevrolet Monte Carlo Coupe 2007',
 73: 'Chevrolet Malibu Sedan 2007',
 74: 'Chevrolet Silverado 1500 Extended Cab 2012',
 75: 'Chevrolet Silverado 1500 Regular Cab 2012',
 76: 'Chrysler Aspen SUV 2009',
 77: 'Chrysler Sebring Convertible 2010',
 78: 'Chrysler Town and Country Minivan 2012',
 79: 'Chrysler 300 SRT-8 2010',
 80: 'Chrysler Crossfire Convertible 2008',
 81: 'Chrysler PT Cruiser Convertible 2008',
 82: 'Daewoo Nubira Wagon 2002',
 83: 'Dodge Caliber Wagon 2012',
 84: 'Dodge Caliber Wagon 2007',
 85: 'Dodge Caravan Minivan 1997',
 86: 'Dodge Ram Pickup 3500 Crew Cab 2010',
 87: 'Dodge Ram Pickup 3500 Quad Cab 2009',
 88: 'Dodge Sprinter Cargo Van 2009',
 89: 'Dodge Journey SUV 2012',
 90: 'Dodge Dakota Crew Cab 2010',
 91: 'Dodge Dakota Club Cab 2007',
 92: 'Dodge Magnum Wagon 2008',
 93: 'Dodge Challenger SRT8 2011',
 94: 'Dodge Durango SUV 2012',
 95: 'Dodge Durango SUV 2007',
 96: 'Dodge Charger Sedan 2012',
 97: 'Dodge Charger SRT-8 2009',
 98: 'Eagle Talon Hatchback 1998',
 99: 'FIAT 500 Abarth 2012',
 100: 'FIAT 500 Convertible 2012',
 101: 'Ferrari FF Coupe 2012',
 102: 'Ferrari California Convertible 2012',
 103: 'Ferrari 458 Italia Convertible 2012',
 104: 'Ferrari 458 Italia Coupe 2012',
 105: 'Fisker Karma Sedan 2012',
 106: 'Ford F-450 Super Duty Crew Cab 2012',
 107: 'Ford Mustang Convertible 2007',
 108: 'Ford Freestar Minivan 2007',
 109: 'Ford Expedition EL SUV 2009',
 110: 'Ford Edge SUV 2012',
 111: 'Ford Ranger SuperCab 2011',
 112: 'Ford GT Coupe 2006',
 113: 'Ford F-150 Regular Cab 2012',
 114: 'Ford F-150 Regular Cab 2007',
 115: 'Ford Focus Sedan 2007',
 116: 'Ford E-Series Wagon Van 2012',
 117: 'Ford Fiesta Sedan 2012',
 118: 'GMC Terrain SUV 2012',
 119: 'GMC Savana Van 2012',
 120: 'GMC Yukon Hybrid SUV 2012',
 121: 'GMC Acadia SUV 2012',
 122: 'GMC Canyon Extended Cab 2012',
 123: 'Geo Metro Convertible 1993',
 124: 'HUMMER H3T Crew Cab 2010',
 125: 'HUMMER H2 SUT Crew Cab 2009',
 126: 'Honda Odyssey Minivan 2012',
 127: 'Honda Odyssey Minivan 2007',
 128: 'Honda Accord Coupe 2012',
 129: 'Honda Accord Sedan 2012',
 130: 'Hyundai Veloster Hatchback 2012',
 131: 'Hyundai Santa Fe SUV 2012',
 132: 'Hyundai Tucson SUV 2012',
 133: 'Hyundai Veracruz SUV 2012',
 134: 'Hyundai Sonata Hybrid Sedan 2012',
 135: 'Hyundai Elantra Sedan 2007',
 136: 'Hyundai Accent Sedan 2012',
 137: 'Hyundai Genesis Sedan 2012',
 138: 'Hyundai Sonata Sedan 2012',
 139: 'Hyundai Elantra Touring Hatchback 2012',
 140: 'Hyundai Azera Sedan 2012',
 141: 'Infiniti G Coupe IPL 2012',
 142: 'Infiniti QX56 SUV 2011',
 143: 'Isuzu Ascender SUV 2008',
 144: 'Jaguar XK XKR 2012',
 145: 'Jeep Patriot SUV 2012',
 146: 'Jeep Wrangler SUV 2012',
 147: 'Jeep Liberty SUV 2012',
 148: 'Jeep Grand Cherokee SUV 2012',
 149: 'Jeep Compass SUV 2012',
 150: 'Lamborghini Reventon Coupe 2008',
 151: 'Lamborghini Aventador Coupe 2012',
 152: 'Lamborghini Gallardo LP 570-4 Superleggera 2012',
 153: 'Lamborghini Diablo Coupe 2001',
 154: 'Land Rover Range Rover SUV 2012',
 155: 'Land Rover LR2 SUV 2012',
 156: 'Lincoln Town Car Sedan 2011',
 157: 'MINI Cooper Roadster Convertible 2012',
 158: 'Maybach Landaulet Convertible 2012',
 159: 'Mazda Tribute SUV 2011',
 160: 'McLaren MP4-12C Coupe 2012',
 161: 'Mercedes-Benz 300-Class Convertible 1993',
 162: 'Mercedes-Benz C-Class Sedan 2012',
 163: 'Mercedes-Benz SL-Class Coupe 2009',
 164: 'Mercedes-Benz E-Class Sedan 2012',
 165: 'Mercedes-Benz S-Class Sedan 2012',
 166: 'Mercedes-Benz Sprinter Van 2012',
 167: 'Mitsubishi Lancer Sedan 2012',
 168: 'Nissan Leaf Hatchback 2012',
 169: 'Nissan NV Passenger Van 2012',
 170: 'Nissan Juke Hatchback 2012',
 171: 'Nissan 240SX Coupe 1998',
 172: 'Plymouth Neon Coupe 1999',
 173: 'Porsche Panamera Sedan 2012',
 174: 'Ram C/V Cargo Van Minivan 2012',
 175: 'Rolls-Royce Phantom Drophead Coupe Convertible 2012',
 176: 'Rolls-Royce Ghost Sedan 2012',
 177: 'Rolls-Royce Phantom Sedan 2012',
 178: 'Scion xD Hatchback 2012',
 179: 'Spyker C8 Convertible 2009',
 180: 'Spyker C8 Coupe 2009',
 181: 'Suzuki Aerio Sedan 2007',
 182: 'Suzuki Kizashi Sedan 2012',
 183: 'Suzuki SX4 Hatchback 2012',
 184: 'Suzuki SX4 Sedan 2012',
 185: 'Tesla Model S Sedan 2012',
 186: 'Toyota Sequoia SUV 2012',
 187: 'Toyota Camry Sedan 2012',
 188: 'Toyota Corolla Sedan 2012',
 189: 'Toyota 4Runner SUV 2012',
 190: 'Volkswagen Golf Hatchback 2012',
 191: 'Volkswagen Golf Hatchback 1991',
 192: 'Volkswagen Beetle Hatchback 2012',
 193: 'Volvo C30 Hatchback 2012',
 194: 'Volvo 240 Sedan 1993',
 195: 'Volvo XC90 SUV 2007',
 196: 'smart fortwo Convertible 2012'}
In [48]:
train_img_dir = os.path.join(car_img_path,'Train Images')
test_img_dir =  os.path.join(car_img_path,'Test Images')
In [49]:
ds_train_ann = pd.read_csv(os.path.join(car_ann_path,'Train Annotations.csv'),names=['Image_Name','Start_x','Start_y','End_x','End_y','Image_class'])
ds_test_ann = pd.read_csv(os.path.join(car_ann_path,'Test Annotation.csv'),names=['Image_Name','Start_x','Start_y','End_x','End_y','Image_class'])
In [50]:
ds_train_ann.drop([ds_train_ann.index[0]],inplace=True)
ds_test_ann.drop([ds_test_ann.index[0]],inplace=True)
In [51]:
ds_train_ann.reset_index(inplace=True,drop=True)
ds_test_ann.reset_index(inplace=True,drop=True)
ds_train_ann.head(5)
Out[51]:
Image_Name Start_x Start_y End_x End_y Image_class
0 00001.jpg 39 116.0 569.0 375.0 14
1 00002.jpg 36 116.0 868.0 587.0 3
2 00003.jpg 85 109.0 601.0 381.0 91
3 00004.jpg 621 393.0 1484.0 1096.0 134
4 00005.jpg 14 36.0 133.0 99.0 106
In [52]:
sns.histplot(ds_train_ann['Image_class'].value_counts())
Out[52]:
<AxesSubplot:xlabel='Image_class', ylabel='Count'>
In [53]:
sns.histplot(ds_test_ann['Image_class'].value_counts())
Out[53]:
<AxesSubplot:xlabel='Image_class', ylabel='Count'>
In [54]:
def load_images_from_folder(folder):
    #file_name = []
    file_path = {}
    count = 0
    for folders in os.listdir(folder):
        #print(folders)
        if folders == '.DS_Store':
            pass
        else:
            foldername = os.path.join(folder,folders)
            #print(foldername)
            for filename in os.listdir(foldername):
                count = count+1
                filepath = os.path.join(foldername,filename)
                #file_name.append(filename)
                file_path[filename]=filepath
                #print(filename)
    return file_path
In [55]:
train_images_path = load_images_from_folder(train_img_dir)
test_images_path = load_images_from_folder(test_img_dir)
ds_train_ann['Image_Path'] = ds_train_ann['Image_Name'].apply(lambda x:train_images_path[x])
ds_train_ann['Car_Name'] = ds_train_ann['Image_class'].apply(lambda x:car_dict[int(x)])
ds_train_ann['Start_x'] = ds_train_ann.Start_x.astype(float)
ds_train_ann['Image_class'] = ds_train_ann.Image_class.astype(int)
ds_train_ann['Bbox'] = ds_train_ann.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
ds_test_ann['Image_Path'] = ds_test_ann['Image_Name'].apply(lambda x:test_images_path[x])
ds_test_ann['Start_x'] = ds_test_ann.Start_x.astype(float)
ds_test_ann['Image_class'] = ds_test_ann.Image_class.astype(int)
ds_test_ann['Car_Name'] = ds_test_ann['Image_class'].apply(lambda x:car_dict[int(x)])
ds_test_ann['Bbox'] = ds_test_ann.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
In [56]:
ds_train_ann
Out[56]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox
0 00001.jpg 39.0 116.0 569.0 375.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [39.0, 116.0, 569.0, 375.0]
1 00002.jpg 36.0 116.0 868.0 587.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [36.0, 116.0, 868.0, 587.0]
2 00003.jpg 85.0 109.0 601.0 381.0 91 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Dakota Club Cab 2007 [85.0, 109.0, 601.0, 381.0]
3 00004.jpg 621.0 393.0 1484.0 1096.0 134 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Hybrid Sedan 2012 [621.0, 393.0, 1484.0, 1096.0]
4 00005.jpg 14.0 36.0 133.0 99.0 106 C:\Users\uprak\capstone\Car Images\Train Image... Ford F-450 Super Duty Crew Cab 2012 [14.0, 36.0, 133.0, 99.0]
... ... ... ... ... ... ... ... ... ...
8139 08140.jpg 3.0 44.0 423.0 336.0 78 C:\Users\uprak\capstone\Car Images\Train Image... Chrysler Town and Country Minivan 2012 [3.0, 44.0, 423.0, 336.0]
8140 08141.jpg 138.0 150.0 706.0 523.0 196 C:\Users\uprak\capstone\Car Images\Train Image... smart fortwo Convertible 2012 [138.0, 150.0, 706.0, 523.0]
8141 08142.jpg 26.0 246.0 660.0 449.0 163 C:\Users\uprak\capstone\Car Images\Train Image... Mercedes-Benz SL-Class Coupe 2009 [26.0, 246.0, 660.0, 449.0]
8142 08143.jpg 78.0 526.0 1489.0 908.0 112 C:\Users\uprak\capstone\Car Images\Train Image... Ford GT Coupe 2006 [78.0, 526.0, 1489.0, 908.0]
8143 08144.jpg 20.0 240.0 862.0 677.0 17 C:\Users\uprak\capstone\Car Images\Train Image... Audi 100 Sedan 1994 [20.0, 240.0, 862.0, 677.0]

8144 rows × 9 columns

In [128]:
df_bb = ds_train_ann[['Image_Name','Start_x','Start_y','End_x','End_y','Car_Name','Image_class','Image_Path']]
In [129]:
pd.plotting.scatter_matrix(df_bb, alpha=0.5)
Out[129]:
array([[<AxesSubplot:xlabel='Start_x', ylabel='Start_x'>,
        <AxesSubplot:xlabel='Start_y', ylabel='Start_x'>,
        <AxesSubplot:xlabel='End_x', ylabel='Start_x'>,
        <AxesSubplot:xlabel='End_y', ylabel='Start_x'>,
        <AxesSubplot:xlabel='Image_class', ylabel='Start_x'>],
       [<AxesSubplot:xlabel='Start_x', ylabel='Start_y'>,
        <AxesSubplot:xlabel='Start_y', ylabel='Start_y'>,
        <AxesSubplot:xlabel='End_x', ylabel='Start_y'>,
        <AxesSubplot:xlabel='End_y', ylabel='Start_y'>,
        <AxesSubplot:xlabel='Image_class', ylabel='Start_y'>],
       [<AxesSubplot:xlabel='Start_x', ylabel='End_x'>,
        <AxesSubplot:xlabel='Start_y', ylabel='End_x'>,
        <AxesSubplot:xlabel='End_x', ylabel='End_x'>,
        <AxesSubplot:xlabel='End_y', ylabel='End_x'>,
        <AxesSubplot:xlabel='Image_class', ylabel='End_x'>],
       [<AxesSubplot:xlabel='Start_x', ylabel='End_y'>,
        <AxesSubplot:xlabel='Start_y', ylabel='End_y'>,
        <AxesSubplot:xlabel='End_x', ylabel='End_y'>,
        <AxesSubplot:xlabel='End_y', ylabel='End_y'>,
        <AxesSubplot:xlabel='Image_class', ylabel='End_y'>],
       [<AxesSubplot:xlabel='Start_x', ylabel='Image_class'>,
        <AxesSubplot:xlabel='Start_y', ylabel='Image_class'>,
        <AxesSubplot:xlabel='End_x', ylabel='Image_class'>,
        <AxesSubplot:xlabel='End_y', ylabel='Image_class'>,
        <AxesSubplot:xlabel='Image_class', ylabel='Image_class'>]],
      dtype=object)
In [130]:
sns.relplot(x='Start_x',y='End_x',data=df_bb)
Out[130]:
<seaborn.axisgrid.FacetGrid at 0x206c0f62d30>
In [131]:
sns.relplot(x='Start_y',y='End_y',data=df_bb)
Out[131]:
<seaborn.axisgrid.FacetGrid at 0x206c0f41cd0>
In [132]:
sns.relplot(x='Start_x',y='Start_y',data=df_bb)
Out[132]:
<seaborn.axisgrid.FacetGrid at 0x206c0f03820>
In [133]:
sns.relplot(x='End_x',y='End_y',data=df_bb)
Out[133]:
<seaborn.axisgrid.FacetGrid at 0x2070aae9b80>
In [173]:
df_bb['width'] = (df_bb['End_x'] - df_bb['Start_x'])
df_bb['height'] = (df_bb['End_y'] - df_bb['Start_y'])
df_bb['area'] = df_bb['height'] * df_bb['width']
In [174]:
df_bb.head(5)
Out[174]:
Image_Name Start_x Start_y End_x End_y Car_Name Image_class Image_Path width height area
0 00001.jpg 39.0 116.0 569.0 375.0 Audi TTS Coupe 2012 14 C:\Users\uprak\capstone\Car Images\Train Image... 530.0 259.0 137270.0
1 00002.jpg 36.0 116.0 868.0 587.0 Acura TL Sedan 2012 3 C:\Users\uprak\capstone\Car Images\Train Image... 832.0 471.0 391872.0
2 00003.jpg 85.0 109.0 601.0 381.0 Dodge Dakota Club Cab 2007 91 C:\Users\uprak\capstone\Car Images\Train Image... 516.0 272.0 140352.0
3 00004.jpg 621.0 393.0 1484.0 1096.0 Hyundai Sonata Hybrid Sedan 2012 134 C:\Users\uprak\capstone\Car Images\Train Image... 863.0 703.0 606689.0
4 00005.jpg 14.0 36.0 133.0 99.0 Ford F-450 Super Duty Crew Cab 2012 106 C:\Users\uprak\capstone\Car Images\Train Image... 119.0 63.0 7497.0
In [175]:
sns.relplot(x='width',y='height',hue='area',palette="ch:r=-.5,l=.75",data=df_bb)
Out[175]:
<seaborn.axisgrid.FacetGrid at 0x2070a44ce20>
In [191]:
df_bb[df_bb['area']<=5000.0]
Out[191]:
Image_Name Start_x Start_y End_x End_y Car_Name Image_class Image_Path width height area
96 00097.jpg 3.0 7.0 76.0 50.0 Chevrolet Corvette Ron Fellows Edition Z06 2007 57 C:\Users\uprak\capstone\Car Images\Train Image... 73.0 43.0 3139.0
215 00216.jpg 21.0 45.0 133.0 88.0 GMC Savana Van 2012 119 C:\Users\uprak\capstone\Car Images\Train Image... 112.0 43.0 4816.0
461 00462.jpg 5.0 8.0 83.0 58.0 AM General Hummer SUV 2000 1 C:\Users\uprak\capstone\Car Images\Train Image... 78.0 50.0 3900.0
1039 01040.jpg 28.0 21.0 105.0 79.0 Dodge Sprinter Cargo Van 2009 88 C:\Users\uprak\capstone\Car Images\Train Image... 77.0 58.0 4466.0
1786 01787.jpg 11.0 26.0 98.0 68.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 87.0 42.0 3654.0
2046 02047.jpg 43.0 50.0 136.0 90.0 Chevrolet Corvette Ron Fellows Edition Z06 2007 57 C:\Users\uprak\capstone\Car Images\Train Image... 93.0 40.0 3720.0
2291 02292.jpg 2.0 15.0 83.0 61.0 Audi 100 Wagon 1994 18 C:\Users\uprak\capstone\Car Images\Train Image... 81.0 46.0 3726.0
2427 02428.jpg 5.0 5.0 92.0 62.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 87.0 57.0 4959.0
3434 03435.jpg 4.0 15.0 87.0 54.0 Audi 100 Wagon 1994 18 C:\Users\uprak\capstone\Car Images\Train Image... 83.0 39.0 3237.0
4046 04047.jpg 12.0 19.0 100.0 54.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 88.0 35.0 3080.0
4589 04590.jpg 83.0 67.0 181.0 114.0 Chevrolet Tahoe Hybrid SUV 2012 62 C:\Users\uprak\capstone\Car Images\Train Image... 98.0 47.0 4606.0
4666 04667.jpg 3.0 9.0 91.0 59.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 88.0 50.0 4400.0
5223 05224.jpg 5.0 14.0 93.0 58.0 AM General Hummer SUV 2000 1 C:\Users\uprak\capstone\Car Images\Train Image... 88.0 44.0 3872.0
5656 05657.jpg 5.0 8.0 91.0 59.0 AM General Hummer SUV 2000 1 C:\Users\uprak\capstone\Car Images\Train Image... 86.0 51.0 4386.0
5803 05804.jpg 4.0 15.0 94.0 63.0 AM General Hummer SUV 2000 1 C:\Users\uprak\capstone\Car Images\Train Image... 90.0 48.0 4320.0
6283 06284.jpg 7.0 41.0 83.0 64.0 Chevrolet Corvette Ron Fellows Edition Z06 2007 57 C:\Users\uprak\capstone\Car Images\Train Image... 76.0 23.0 1748.0
6361 06362.jpg 2.0 19.0 104.0 60.0 Audi 100 Wagon 1994 18 C:\Users\uprak\capstone\Car Images\Train Image... 102.0 41.0 4182.0
6483 06484.jpg 1.0 11.0 92.0 47.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 91.0 36.0 3276.0
7346 07347.jpg 6.0 12.0 86.0 49.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 80.0 37.0 2960.0
7468 07469.jpg 3.0 13.0 76.0 52.0 Chevrolet Corvette Ron Fellows Edition Z06 2007 57 C:\Users\uprak\capstone\Car Images\Train Image... 73.0 39.0 2847.0
8086 08087.jpg 5.0 14.0 84.0 55.0 Geo Metro Convertible 1993 123 C:\Users\uprak\capstone\Car Images\Train Image... 79.0 41.0 3239.0
In [179]:
sns.distplot(df_bb['area'])
Out[179]:
<AxesSubplot:xlabel='area', ylabel='Density'>
In [181]:
df_bb[df_bb['area']>=4000000]
Out[181]:
Image_Name Start_x Start_y End_x End_y Car_Name Image_class Image_Path width height area
76 00077.jpg 25.0 25.0 3502.0 2017.0 Ford Edge SUV 2012 110 C:\Users\uprak\capstone\Car Images\Train Image... 3477.0 1992.0 6926184.0
229 00230.jpg 223.0 542.0 3692.0 2478.0 Suzuki SX4 Hatchback 2012 183 C:\Users\uprak\capstone\Car Images\Train Image... 3469.0 1936.0 6715984.0
258 00259.jpg 75.0 254.0 2514.0 1929.0 Toyota Camry Sedan 2012 187 C:\Users\uprak\capstone\Car Images\Train Image... 2439.0 1675.0 4085325.0
345 00346.jpg 112.0 21.0 2859.0 1928.0 Chevrolet Sonic Sedan 2012 63 C:\Users\uprak\capstone\Car Images\Train Image... 2747.0 1907.0 5238529.0
1032 01033.jpg 49.0 279.0 2791.0 1870.0 Chevrolet Impala Sedan 2007 61 C:\Users\uprak\capstone\Car Images\Train Image... 2742.0 1591.0 4362522.0
1062 01063.jpg 351.0 286.0 3753.0 2243.0 Chrysler Town and Country Minivan 2012 78 C:\Users\uprak\capstone\Car Images\Train Image... 3402.0 1957.0 6657714.0
1426 01427.jpg 347.0 476.0 4243.0 2904.0 Jeep Compass SUV 2012 149 C:\Users\uprak\capstone\Car Images\Train Image... 3896.0 2428.0 9459488.0
2068 02069.jpg 1.0 550.0 2320.0 2506.0 Nissan Leaf Hatchback 2012 168 C:\Users\uprak\capstone\Car Images\Train Image... 2319.0 1956.0 4535964.0
2078 02079.jpg 45.0 181.0 2582.0 1848.0 Chevrolet Silverado 1500 Regular Cab 2012 75 C:\Users\uprak\capstone\Car Images\Train Image... 2537.0 1667.0 4229179.0
2086 02087.jpg 19.0 533.0 3220.0 1953.0 BMW 3 Series Wagon 2012 30 C:\Users\uprak\capstone\Car Images\Train Image... 3201.0 1420.0 4545420.0
2142 02143.jpg 8.0 76.0 3640.0 2675.0 Plymouth Neon Coupe 1999 172 C:\Users\uprak\capstone\Car Images\Train Image... 3632.0 2599.0 9439568.0
2363 02364.jpg 373.0 56.0 2663.0 1915.0 Acura TL Sedan 2012 3 C:\Users\uprak\capstone\Car Images\Train Image... 2290.0 1859.0 4257110.0
2995 02996.jpg 174.0 271.0 3826.0 2919.0 Jeep Compass SUV 2012 149 C:\Users\uprak\capstone\Car Images\Train Image... 3652.0 2648.0 9670496.0
3184 03185.jpg 673.0 981.0 4278.0 2553.0 Chevrolet Corvette ZR1 2012 56 C:\Users\uprak\capstone\Car Images\Train Image... 3605.0 1572.0 5667060.0
3229 03230.jpg 36.0 122.0 3433.0 1518.0 Mitsubishi Lancer Sedan 2012 167 C:\Users\uprak\capstone\Car Images\Train Image... 3397.0 1396.0 4742212.0
3488 03489.jpg 1.0 329.0 3054.0 2242.0 Jeep Compass SUV 2012 149 C:\Users\uprak\capstone\Car Images\Train Image... 3053.0 1913.0 5840389.0
4103 04104.jpg 48.0 74.0 2608.0 1640.0 Chevrolet Silverado 1500 Classic Extended Cab ... 70 C:\Users\uprak\capstone\Car Images\Train Image... 2560.0 1566.0 4008960.0
4136 04137.jpg 199.0 482.0 3385.0 2465.0 Chevrolet Traverse SUV 2012 58 C:\Users\uprak\capstone\Car Images\Train Image... 3186.0 1983.0 6317838.0
4302 04303.jpg 192.0 581.0 3976.0 1791.0 Mercedes-Benz E-Class Sedan 2012 164 C:\Users\uprak\capstone\Car Images\Train Image... 3784.0 1210.0 4578640.0
4397 04398.jpg 721.0 172.0 3678.0 1982.0 Rolls-Royce Ghost Sedan 2012 176 C:\Users\uprak\capstone\Car Images\Train Image... 2957.0 1810.0 5352170.0
5224 05225.jpg 108.0 103.0 2743.0 1758.0 Scion xD Hatchback 2012 178 C:\Users\uprak\capstone\Car Images\Train Image... 2635.0 1655.0 4360925.0
5436 05437.jpg 96.0 416.0 4252.0 2792.0 Chevrolet Impala Sedan 2007 61 C:\Users\uprak\capstone\Car Images\Train Image... 4156.0 2376.0 9874656.0
5536 05537.jpg 667.0 274.0 3686.0 2221.0 Rolls-Royce Ghost Sedan 2012 176 C:\Users\uprak\capstone\Car Images\Train Image... 3019.0 1947.0 5877993.0
5554 05555.jpg 144.0 643.0 3656.0 2288.0 Chevrolet Corvette Ron Fellows Edition Z06 2007 57 C:\Users\uprak\capstone\Car Images\Train Image... 3512.0 1645.0 5777240.0
5944 05945.jpg 1648.0 1508.0 5205.0 3389.0 Chevrolet Sonic Sedan 2012 63 C:\Users\uprak\capstone\Car Images\Train Image... 3557.0 1881.0 6690717.0
6130 06131.jpg 271.0 440.0 3392.0 1955.0 Ford Edge SUV 2012 110 C:\Users\uprak\capstone\Car Images\Train Image... 3121.0 1515.0 4728315.0
6451 06452.jpg 645.0 281.0 3464.0 2572.0 Jeep Patriot SUV 2012 145 C:\Users\uprak\capstone\Car Images\Train Image... 2819.0 2291.0 6458329.0
7192 07193.jpg 542.0 749.0 3860.0 2657.0 BMW M6 Convertible 2010 36 C:\Users\uprak\capstone\Car Images\Train Image... 3318.0 1908.0 6330744.0
7329 07330.jpg 143.0 541.0 3812.0 2094.0 Hyundai Azera Sedan 2012 140 C:\Users\uprak\capstone\Car Images\Train Image... 3669.0 1553.0 5697957.0
7385 07386.jpg 178.0 790.0 3061.0 2333.0 Aston Martin Virage Convertible 2012 10 C:\Users\uprak\capstone\Car Images\Train Image... 2883.0 1543.0 4448469.0
7539 07540.jpg 417.0 497.0 2981.0 2088.0 Volkswagen Golf Hatchback 1991 191 C:\Users\uprak\capstone\Car Images\Train Image... 2564.0 1591.0 4079324.0
In [183]:
df_visualize = df_bb[df_bb['area']>=4000000]
df_visualize['Bbox'] = df_visualize.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
image_paths = list(df_visualize['Image_Path']) 
images = [read_image(p) for p in image_paths]
#images = np.array(df_visualize['Images'])
image_name = list(df_visualize['Car_Name'])
bbox_list = list(df_visualize['Bbox'])
image = visualize(images,bbox_list,image_name)
fig = plt.figure(figsize=(20, 20))
columns = 4
rows = 8

pics = []
for i in range(columns*rows-1):
    pics.append(fig.add_subplot(rows, columns, i+1,title=image_name[i]))
    plt.imshow(image[i])

plt.show()
In [192]:
df_visualize = df_bb[df_bb['area']<=5000.0]
df_visualize['Bbox'] = df_visualize.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
image_paths = list(df_visualize['Image_Path']) 
images = [read_image(p) for p in image_paths]
#images = np.array(df_visualize['Images'])
image_name = list(df_visualize['Car_Name'])
bbox_list = list(df_visualize['Bbox'])
image = visualize(images,bbox_list,image_name)
fig = plt.figure(figsize=(20, 20))
columns = 4
rows = 6

pics = []
for i in range(columns*rows-3):
    pics.append(fig.add_subplot(rows, columns, i+1,title=image_name[i]))
    plt.imshow(image[i])

plt.show()
In [19]:
def get_image_dims(df,col):
    height = []
    width = []
    for img_path in df[col].values:
        img = cv.imread(img_path)
        height.append(int(img.shape[0]))
        width.append(int(img.shape[1]))
    return height,width

#ds_train_ann['Orig_img_height'],ds_train_ann['Orig_img_width'] =  get_image_dims(ds_train_ann,'Image_Path')
#ds_test_ann['Orig_img_height'],ds_test_ann['Orig_img_width'] =  get_image_dims(ds_test_ann,'Image_Path')
In [20]:
ds_train_ann
Out[20]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox
0 00001.jpg 39.0 116.0 569.0 375.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [39.0, 116.0, 569.0, 375.0]
1 00002.jpg 36.0 116.0 868.0 587.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [36.0, 116.0, 868.0, 587.0]
2 00003.jpg 85.0 109.0 601.0 381.0 91 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Dakota Club Cab 2007 [85.0, 109.0, 601.0, 381.0]
3 00004.jpg 621.0 393.0 1484.0 1096.0 134 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Hybrid Sedan 2012 [621.0, 393.0, 1484.0, 1096.0]
4 00005.jpg 14.0 36.0 133.0 99.0 106 C:\Users\uprak\capstone\Car Images\Train Image... Ford F-450 Super Duty Crew Cab 2012 [14.0, 36.0, 133.0, 99.0]
... ... ... ... ... ... ... ... ... ...
8139 08140.jpg 3.0 44.0 423.0 336.0 78 C:\Users\uprak\capstone\Car Images\Train Image... Chrysler Town and Country Minivan 2012 [3.0, 44.0, 423.0, 336.0]
8140 08141.jpg 138.0 150.0 706.0 523.0 196 C:\Users\uprak\capstone\Car Images\Train Image... smart fortwo Convertible 2012 [138.0, 150.0, 706.0, 523.0]
8141 08142.jpg 26.0 246.0 660.0 449.0 163 C:\Users\uprak\capstone\Car Images\Train Image... Mercedes-Benz SL-Class Coupe 2009 [26.0, 246.0, 660.0, 449.0]
8142 08143.jpg 78.0 526.0 1489.0 908.0 112 C:\Users\uprak\capstone\Car Images\Train Image... Ford GT Coupe 2006 [78.0, 526.0, 1489.0, 908.0]
8143 08144.jpg 20.0 240.0 862.0 677.0 17 C:\Users\uprak\capstone\Car Images\Train Image... Audi 100 Sedan 1994 [20.0, 240.0, 862.0, 677.0]

8144 rows × 9 columns

In [21]:
#ds_train_ann.to_csv("train1.csv", index=False)
In [22]:
#ds_test_ann.to_csv("test1.csv",index=False)
In [23]:
df_train = pd.read_csv("train1.csv")
df_test = pd.read_csv("test1.csv")
In [24]:
df_train.sample(10)
Out[24]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images
4815 04816.jpg 79.0 84.0 528.0 327.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [79.0, 84.0, 528.0, 327.0] 404 606 1.500000 449.0 243.0 1.847737 244824 109107.0 [[[193 171 111]\n [200 176 114]\n [198 173 1...
3264 03265.jpg 33.0 150.0 619.0 409.0 6 C:\Users\uprak\capstone\Car Images\Train Image... Acura Integra Type R 2001 [33.0, 150.0, 619.0, 409.0] 480 640 1.333333 586.0 259.0 2.262548 307200 151774.0 [[[ 51 11 12]\n [ 53 13 14]\n [ 56 16 ...
4342 04343.jpg 143.0 310.0 915.0 656.0 112 C:\Users\uprak\capstone\Car Images\Train Image... Ford GT Coupe 2006 [143.0, 310.0, 915.0, 656.0] 768 1024 1.333333 772.0 346.0 2.231214 786432 267112.0 [[[156 175 233]\n [154 173 231]\n [152 171 2...
6332 06333.jpg 80.0 81.0 575.0 393.0 85 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Caravan Minivan 1997 [80.0, 81.0, 575.0, 393.0] 480 640 1.333333 495.0 312.0 1.586538 307200 154440.0 [[[ 73 83 75]\n [ 63 73 65]\n [ 47 57 ...
3320 03321.jpg 8.0 59.0 498.0 266.0 16 C:\Users\uprak\capstone\Car Images\Train Image... Audi V8 Sedan 1994 [8.0, 59.0, 498.0, 266.0] 315 500 1.587302 490.0 207.0 2.367150 157500 101430.0 [[[122 90 65]\n [120 88 63]\n [120 88 ...
6880 06881.jpg 28.0 75.0 618.0 290.0 84 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Caliber Wagon 2007 [28.0, 75.0, 618.0, 290.0] 360 640 1.777778 590.0 215.0 2.744186 230400 126850.0 [[[ 1 1 1]\n [ 1 1 1]\n [ 1 1 ...
1824 01825.jpg 79.0 49.0 555.0 421.0 35 C:\Users\uprak\capstone\Car Images\Train Image... BMW M5 Sedan 2010 [79.0, 49.0, 555.0, 421.0] 480 640 1.333333 476.0 372.0 1.279570 307200 177072.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
7349 07350.jpg 24.0 48.0 646.0 337.0 128 C:\Users\uprak\capstone\Car Images\Train Image... Honda Accord Coupe 2012 [24.0, 48.0, 646.0, 337.0] 370 660 1.783784 622.0 289.0 2.152249 244200 179758.0 [[[ 71 78 84]\n [ 72 79 85]\n [ 73 80 ...
2237 02238.jpg 68.0 135.0 389.0 282.0 138 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Sedan 2012 [68.0, 135.0, 389.0, 282.0] 360 540 1.500000 321.0 147.0 2.183673 194400 47187.0 [[[125 106 110]\n [125 106 110]\n [126 107 1...
4571 04572.jpg 3.0 122.0 550.0 333.0 82 C:\Users\uprak\capstone\Car Images\Train Image... Daewoo Nubira Wagon 2002 [3.0, 122.0, 550.0, 333.0] 415 554 1.334940 547.0 211.0 2.592417 229910 115417.0 [[[224 229 248]\n [226 232 248]\n [226 229 2...
In [25]:
#df_train['Img_AR'] = df_train['Orig_img_width']/df_train['Orig_img_height']
#df_train['bb_width'] = df_train['End_x'] - df_train['Start_x']
#df_train['bb_height'] = df_train['End_y'] - df_train['Start_y']
#df_train['bb_AR'] = df_train['bb_width']/df_train['bb_height']
In [26]:
#df_test['Img_AR'] = df_test['Orig_img_width']/df_test['Orig_img_height']
#df_test['bb_width'] = df_test['End_x'] - df_test['Start_x']
#df_test['bb_height'] = df_test['End_y'] - df_test['Start_y']
#df_test['bb_AR'] = df_test['bb_width']/df_test['bb_height']
In [27]:
df_train.sample(5)
Out[27]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images
3817 03818.jpg 96.0 82.0 709.0 468.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [96.0, 82.0, 709.0, 468.0] 600 800 1.333333 613.0 386.0 1.588083 480000 236618.0 [[[132 163 96]\n [124 155 88]\n [115 146 ...
1844 01845.jpg 65.0 65.0 268.0 180.0 162 C:\Users\uprak\capstone\Car Images\Train Image... Mercedes-Benz C-Class Sedan 2012 [65.0, 65.0, 268.0, 180.0] 225 300 1.333333 203.0 115.0 1.765217 67500 23345.0 [[[254 254 254]\n [255 255 255]\n [255 255 2...
8043 08044.jpg 32.0 121.0 651.0 410.0 10 C:\Users\uprak\capstone\Car Images\Train Image... Aston Martin Virage Convertible 2012 [32.0, 121.0, 651.0, 410.0] 455 680 1.494505 619.0 289.0 2.141869 309400 178891.0 [[[ 15 14 10]\n [ 17 16 12]\n [ 16 14 ...
6792 06793.jpg 17.0 71.0 227.0 175.0 21 C:\Users\uprak\capstone\Car Images\Train Image... Audi S5 Convertible 2012 [17.0, 71.0, 227.0, 175.0] 194 259 1.335052 210.0 104.0 2.019231 50246 21840.0 [[[189 176 183]\n [189 176 183]\n [189 176 1...
5715 05716.jpg 110.0 64.0 1036.0 677.0 89 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Journey SUV 2012 [110.0, 64.0, 1036.0, 677.0] 768 1149 1.496094 926.0 613.0 1.510604 882432 567638.0 [[[255 242 235]\n [255 243 241]\n [255 248 2...
In [28]:
#df_train['img_area'] = df_train['Orig_img_width']*df_train['Orig_img_height']
#df_train['bb_area'] = df_train['bb_width'] * df_train['bb_height']
#df_test['img_area'] = df_test['Orig_img_width']*df_test['Orig_img_height']
#df_test['bb_area'] = df_test['bb_width'] * df_test['bb_height']
In [29]:
#df_train.sample(5)
In [30]:
cols = list(df_train.columns)
cols
Out[30]:
['Image_Name',
 'Start_x',
 'Start_y',
 'End_x',
 'End_y',
 'Image_class',
 'Image_Path',
 'Car_Name',
 'Bbox',
 'Orig_img_height',
 'Orig_img_width',
 'Img_AR',
 'bb_width',
 'bb_height',
 'bb_AR',
 'img_area',
 'bb_area',
 'Images']
In [31]:
df_test.columns
Out[31]:
Index(['Image_Name', 'Start_x', 'Start_y', 'End_x', 'End_y', 'Image_class',
       'Image_Path', 'Car_Name', 'Bbox', 'Orig_img_height', 'Orig_img_width',
       'Images'],
      dtype='object')
In [32]:
df_test
Out[32]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Images
0 00001.jpg 30.0 52.0 246.0 147.0 181 C:\Users\uprak\capstone\Car Images\Test Images... Suzuki Aerio Sedan 2007 [30.0, 52.0, 246.0, 147.0] 182 276 [[[254 254 254]\n [254 254 254]\n [254 254 2...
1 00002.jpg 100.0 19.0 576.0 203.0 103 C:\Users\uprak\capstone\Car Images\Test Images... Ferrari 458 Italia Convertible 2012 [100.0, 19.0, 576.0, 203.0] 360 640 [[[126 122 119]\n [124 120 117]\n [121 117 1...
2 00003.jpg 51.0 105.0 968.0 659.0 145 C:\Users\uprak\capstone\Car Images\Test Images... Jeep Patriot SUV 2012 [51.0, 105.0, 968.0, 659.0] 741 1024 [[[161 160 166]\n [159 158 164]\n [157 156 1...
3 00004.jpg 67.0 84.0 581.0 407.0 187 C:\Users\uprak\capstone\Car Images\Test Images... Toyota Camry Sedan 2012 [67.0, 84.0, 581.0, 407.0] 480 640 [[[ 21 21 21]\n [ 24 24 24]\n [ 22 22 ...
4 00005.jpg 140.0 151.0 593.0 339.0 185 C:\Users\uprak\capstone\Car Images\Test Images... Tesla Model S Sedan 2012 [140.0, 151.0, 593.0, 339.0] 373 600 [[[254 254 242]\n [254 254 242]\n [254 254 2...
... ... ... ... ... ... ... ... ... ... ... ... ...
8036 08037.jpg 49.0 57.0 1169.0 669.0 63 C:\Users\uprak\capstone\Car Images\Test Images... Chevrolet Sonic Sedan 2012 [49.0, 57.0, 1169.0, 669.0] 800 1200 [[[ 93 77 77]\n [ 92 76 76]\n [ 91 75 ...
8037 08038.jpg 23.0 18.0 640.0 459.0 16 C:\Users\uprak\capstone\Car Images\Test Images... Audi V8 Sedan 1994 [23.0, 18.0, 640.0, 459.0] 480 640 [[[ 54 51 42]\n [ 50 47 38]\n [ 51 48 ...
8038 08039.jpg 33.0 27.0 602.0 252.0 17 C:\Users\uprak\capstone\Car Images\Test Images... Audi 100 Sedan 1994 [33.0, 27.0, 602.0, 252.0] 303 635 [[[122 83 54]\n [112 75 46]\n [120 83 ...
8039 08040.jpg 33.0 142.0 521.0 376.0 38 C:\Users\uprak\capstone\Car Images\Test Images... BMW Z4 Convertible 2012 [33.0, 142.0, 521.0, 376.0] 427 640 [[[ 74 61 44]\n [ 74 61 44]\n [ 74 61 ...
8040 08041.jpg 77.0 73.0 506.0 380.0 32 C:\Users\uprak\capstone\Car Images\Test Images... BMW X5 SUV 2007 [77.0, 73.0, 506.0, 380.0] 426 640 [[[ 39 39 31]\n [ 71 71 63]\n [ 45 45 ...

8041 rows × 12 columns

In [33]:
df_test.sample(5)
Out[33]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Images
4093 04094.jpg 229.0 254.0 1473.0 730.0 31 C:\Users\uprak\capstone\Car Images\Test Images... BMW 6 Series Convertible 2007 [229.0, 254.0, 1473.0, 730.0] 1050 1680 [[[ 9 4 1]\n [ 8 4 5]\n [ 0 11 ...
2181 02182.jpg 26.0 132.0 1009.0 675.0 6 C:\Users\uprak\capstone\Car Images\Test Images... Acura Integra Type R 2001 [26.0, 132.0, 1009.0, 675.0] 768 1024 [[[ 15 22 14]\n [ 16 23 15]\n [ 17 24 ...
3390 03391.jpg 76.0 155.0 448.0 275.0 153 C:\Users\uprak\capstone\Car Images\Test Images... Lamborghini Diablo Coupe 2001 [76.0, 155.0, 448.0, 275.0] 333 500 [[[ 4 2 3]\n [ 1 0 0]\n [ 1 0 ...
422 00423.jpg 51.0 281.0 965.0 686.0 34 C:\Users\uprak\capstone\Car Images\Test Images... BMW M3 Coupe 2012 [51.0, 281.0, 965.0, 686.0] 768 1024 [[[202 202 212]\n [206 209 218]\n [202 204 2...
5978 05979.jpg 62.0 59.0 359.0 301.0 139 C:\Users\uprak\capstone\Car Images\Test Images... Hyundai Elantra Touring Hatchback 2012 [62.0, 59.0, 359.0, 301.0] 360 424 [[[255 255 255]\n [255 255 255]\n [255 255 2...
In [34]:
df_group = df_train.groupby(by=['Image_class']).mean()
In [35]:
df_group
Out[35]:
Start_x Start_y End_x End_y Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area
Image_class
1 33.355556 67.333333 449.688889 306.000000 338.777778 491.133333 1.488229 416.333333 238.666667 1.733447 339422.000000 185725.533333
2 69.718750 113.250000 633.250000 380.281250 444.343750 694.250000 1.626136 563.531250 267.031250 2.193562 421385.937500 192106.718750
3 63.953488 99.046512 657.558140 419.860465 482.976744 714.046512 1.490232 593.604651 320.813953 1.999234 510487.860465 294358.186047
4 50.380952 131.261905 673.285714 435.357143 505.428571 720.023810 1.427891 622.904762 304.095238 2.079940 416349.880952 211456.714286
5 47.195122 95.853659 486.926829 313.414634 370.536585 529.878049 1.467877 439.731707 217.560976 2.016969 346084.951220 169860.097561
... ... ... ... ... ... ... ... ... ... ... ... ...
192 116.395349 163.162791 912.906977 600.930233 699.348837 1021.767442 1.480585 796.511628 437.767442 1.924556 825807.209302 433913.860465
193 83.833333 113.166667 693.047619 454.404762 517.261905 778.547619 1.541277 609.214286 341.238095 1.873040 487950.666667 241788.619048
194 84.391304 123.260870 787.304348 485.195652 566.913043 837.391304 1.539087 702.913043 361.934783 2.088240 537252.239130 285973.065217
195 54.674419 74.581395 550.883721 367.627907 409.813953 595.023256 1.478335 496.209302 293.046512 1.713090 288334.674419 171179.627907
196 62.750000 70.550000 553.850000 391.575000 426.250000 623.525000 1.462446 491.100000 321.025000 1.526207 373605.325000 218841.350000

196 rows × 12 columns

In [36]:
df_group.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 196 entries, 1 to 196
Data columns (total 12 columns):
 #   Column           Non-Null Count  Dtype  
---  ------           --------------  -----  
 0   Start_x          196 non-null    float64
 1   Start_y          196 non-null    float64
 2   End_x            196 non-null    float64
 3   End_y            196 non-null    float64
 4   Orig_img_height  196 non-null    float64
 5   Orig_img_width   196 non-null    float64
 6   Img_AR           196 non-null    float64
 7   bb_width         196 non-null    float64
 8   bb_height        196 non-null    float64
 9   bb_AR            196 non-null    float64
 10  img_area         196 non-null    float64
 11  bb_area          196 non-null    float64
dtypes: float64(12)
memory usage: 19.9 KB
In [37]:
df_group.describe().transpose()
Out[37]:
count mean std min 25% 50% 75% max
Start_x 196.0 64.969904 28.874050 22.813953 42.609226 57.319580 77.561163 1.846591e+02
Start_y 196.0 108.318143 37.164405 39.025000 83.025877 101.064857 127.836051 2.608182e+02
End_x 196.0 634.928421 172.482091 317.275000 514.339039 609.108935 701.699275 1.431727e+03
End_y 196.0 414.032831 117.179856 229.000000 333.580952 391.145105 464.607919 9.897955e+02
Orig_img_height 196.0 479.946622 135.123858 252.225000 386.267970 453.641026 529.396959 1.140000e+03
Orig_img_width 196.0 696.466378 194.958120 349.500000 562.917930 665.098930 771.282998 1.579250e+03
Img_AR 196.0 1.475325 0.062479 1.340796 1.427664 1.474263 1.516778 1.694495e+00
bb_width 196.0 569.958517 150.505356 285.575000 467.374492 549.422025 627.529194 1.247068e+03
bb_height 196.0 305.714688 90.111618 169.750000 247.029167 286.353309 338.530454 7.435682e+02
bb_AR 196.0 1.953022 0.200483 1.374284 1.817158 1.943708 2.089676 2.441710e+00
img_area 196.0 471668.232939 282418.224726 130582.125000 298792.795094 402777.703523 543557.405098 1.922553e+06
bb_area 196.0 246246.701083 153703.305672 68165.500000 150833.660714 203522.201252 289702.911364 1.115018e+06
In [38]:
df_train.head(10)
Out[38]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images
0 00001.jpg 39.0 116.0 569.0 375.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [39.0, 116.0, 569.0, 375.0] 400 600 1.500000 530.0 259.0 2.046332 240000 137270.0 [[[101 119 123]\n [ 98 116 120]\n [ 93 111 1...
1 00002.jpg 36.0 116.0 868.0 587.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [36.0, 116.0, 868.0, 587.0] 675 900 1.333333 832.0 471.0 1.766454 607500 391872.0 [[[164 169 175]\n [166 171 177]\n [169 174 1...
2 00003.jpg 85.0 109.0 601.0 381.0 91 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Dakota Club Cab 2007 [85.0, 109.0, 601.0, 381.0] 480 640 1.333333 516.0 272.0 1.897059 307200 140352.0 [[[231 231 231]\n [244 244 244]\n [234 234 2...
3 00004.jpg 621.0 393.0 1484.0 1096.0 134 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Hybrid Sedan 2012 [621.0, 393.0, 1484.0, 1096.0] 1386 2100 1.515152 863.0 703.0 1.227596 2910600 606689.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
4 00005.jpg 14.0 36.0 133.0 99.0 106 C:\Users\uprak\capstone\Car Images\Train Image... Ford F-450 Super Duty Crew Cab 2012 [14.0, 36.0, 133.0, 99.0] 108 144 1.333333 119.0 63.0 1.888889 15552 7497.0 [[[136 167 196]\n [136 167 196]\n [138 169 1...
5 00006.jpg 259.0 289.0 515.0 416.0 123 C:\Users\uprak\capstone\Car Images\Train Image... Geo Metro Convertible 1993 [259.0, 289.0, 515.0, 416.0] 431 575 1.334107 256.0 127.0 2.015748 247825 32512.0 [[[161 183 241]\n [161 183 241]\n [161 183 2...
6 00007.jpg 88.0 80.0 541.0 397.0 89 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Journey SUV 2012 [88.0, 80.0, 541.0, 397.0] 480 640 1.333333 453.0 317.0 1.429022 307200 143601.0 [[[ 40 44 47]\n [ 40 44 47]\n [ 40 44 ...
7 00008.jpg 73.0 79.0 591.0 410.0 96 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Charger Sedan 2012 [73.0, 79.0, 591.0, 410.0] 480 639 1.331250 518.0 331.0 1.564955 306720 171458.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
8 00009.jpg 20.0 126.0 1269.0 771.0 167 C:\Users\uprak\capstone\Car Images\Train Image... Mitsubishi Lancer Sedan 2012 [20.0, 126.0, 1269.0, 771.0] 843 1280 1.518387 1249.0 645.0 1.936434 1079040 805605.0 [[[103 106 99]\n [ 81 76 72]\n [115 104 1...
9 00010.jpg 21.0 110.0 623.0 367.0 58 C:\Users\uprak\capstone\Car Images\Train Image... Chevrolet Traverse SUV 2012 [21.0, 110.0, 623.0, 367.0] 480 640 1.333333 602.0 257.0 2.342412 307200 154714.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
In [39]:
df_train.describe().transpose()
Out[39]:
count mean std min 25% 50% 75% max
Start_x 8144.0 64.906803 82.198684 1.000000 19.000000 39.000000 79.000000 1.648000e+03
Start_y 8144.0 108.661223 104.551635 1.000000 42.000000 80.000000 138.250000 1.508000e+03
End_x 8144.0 638.208620 410.776734 76.000000 392.000000 569.000000 746.250000 5.205000e+03
End_y 8144.0 416.431606 273.786000 47.000000 248.000000 360.000000 477.000000 3.389000e+03
Image_class 8144.0 98.979371 56.503148 1.000000 50.000000 99.000000 148.000000 1.960000e+02
Orig_img_height 8144.0 482.771979 317.580021 57.000000 290.000000 424.000000 540.000000 3.744000e+03
Orig_img_width 8144.0 699.793099 450.922445 78.000000 429.000000 636.500000 800.000000 5.616000e+03
Img_AR 8144.0 1.473561 0.232987 0.593066 1.333333 1.348485 1.519757 3.887538e+00
bb_width 8144.0 573.301817 375.675050 73.000000 338.000000 504.000000 659.000000 4.156000e+03
bb_height 8144.0 307.770383 213.463238 23.000000 173.000000 260.000000 370.000000 2.648000e+03
bb_AR 8144.0 1.951161 0.488941 0.559322 1.593023 1.867606 2.249385 3.975000e+00
img_area 8144.0 477026.917117 890130.689330 4524.000000 121240.000000 265752.000000 432363.750000 2.102630e+07
bb_area 8144.0 249462.668713 494907.035699 1748.000000 59735.000000 130497.000000 237495.750000 9.874656e+06
In [40]:
df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8144 entries, 0 to 8143
Data columns (total 18 columns):
 #   Column           Non-Null Count  Dtype  
---  ------           --------------  -----  
 0   Image_Name       8144 non-null   object 
 1   Start_x          8144 non-null   float64
 2   Start_y          8144 non-null   float64
 3   End_x            8144 non-null   float64
 4   End_y            8144 non-null   float64
 5   Image_class      8144 non-null   int64  
 6   Image_Path       8144 non-null   object 
 7   Car_Name         8144 non-null   object 
 8   Bbox             8144 non-null   object 
 9   Orig_img_height  8144 non-null   int64  
 10  Orig_img_width   8144 non-null   int64  
 11  Img_AR           8144 non-null   float64
 12  bb_width         8144 non-null   float64
 13  bb_height        8144 non-null   float64
 14  bb_AR            8144 non-null   float64
 15  img_area         8144 non-null   int64  
 16  bb_area          8144 non-null   float64
 17  Images           8144 non-null   object 
dtypes: float64(9), int64(4), object(5)
memory usage: 1.1+ MB
In [41]:
df_train[df_train['Orig_img_height'].isna()]
Out[41]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images
In [113]:
def read_image(path):
    im = cv.imread(str(path))
    return cv.cvtColor(im, cv.COLOR_BGR2RGB)

#image_paths = list(df_train['Image_Path']) 
#images = [read_image(p) for p in image_paths]
In [43]:
#df_train['Images'] = images
In [44]:
#df_train.to_csv("train1.csv", index=False)
In [45]:
#image_paths = list(df_test['Image_Path']) 
#images = [read_image(p) for p in image_paths]
In [46]:
#df_test['Images'] = images
In [47]:
#df_test.to_csv("test1.csv", index=False)
In [172]:
BOX_COLOR = (255, 0, 0) # Red
TEXT_COLOR = (255, 255, 255) # White


def visualize_bbox(img, bbox, class_name, color=BOX_COLOR, thickness=2):
    """Visualizes a single bounding box on the image"""
    x_min, y_min, x_max, y_max = map(int,bbox)
    cv.rectangle(img, (x_min, y_min), (x_max, y_max), color=BOX_COLOR, thickness=thickness)
    
    ((text_width, text_height), _) = cv.getTextSize(class_name, cv.FONT_HERSHEY_SIMPLEX, 0.35, 1)    
    cv.rectangle(img, (x_min, (y_min - int(0.3 * text_height))), (x_min + text_width, y_min), BOX_COLOR, -1)
    cv.putText(
        img,
        text=class_name,
        org=(x_min, y_min - int(0.3 * text_height)),
        fontFace=cv.FONT_HERSHEY_SIMPLEX,
        fontScale=0.35, 
        color=TEXT_COLOR, 
        lineType=cv.LINE_AA,
    )
    return img


def visualize(image_list, bbox_list, class_list):
    count = 0
    image_final = []
    for images in image_list:
        #print(img)
        bbox = bbox_list[count]
        #print(bbox)
        class_name = class_list[count]
        #print(class_name)
        img = visualize_bbox(images, bbox, class_name)
        image_final.append(img)
        count = count + 1
    return image_final
In [49]:
df_train['Bbox'] = df_train.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
df_train
Out[49]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images
0 00001.jpg 39.0 116.0 569.0 375.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [39.0, 116.0, 569.0, 375.0] 400 600 1.500000 530.0 259.0 2.046332 240000 137270.0 [[[101 119 123]\n [ 98 116 120]\n [ 93 111 1...
1 00002.jpg 36.0 116.0 868.0 587.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [36.0, 116.0, 868.0, 587.0] 675 900 1.333333 832.0 471.0 1.766454 607500 391872.0 [[[164 169 175]\n [166 171 177]\n [169 174 1...
2 00003.jpg 85.0 109.0 601.0 381.0 91 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Dakota Club Cab 2007 [85.0, 109.0, 601.0, 381.0] 480 640 1.333333 516.0 272.0 1.897059 307200 140352.0 [[[231 231 231]\n [244 244 244]\n [234 234 2...
3 00004.jpg 621.0 393.0 1484.0 1096.0 134 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Hybrid Sedan 2012 [621.0, 393.0, 1484.0, 1096.0] 1386 2100 1.515152 863.0 703.0 1.227596 2910600 606689.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
4 00005.jpg 14.0 36.0 133.0 99.0 106 C:\Users\uprak\capstone\Car Images\Train Image... Ford F-450 Super Duty Crew Cab 2012 [14.0, 36.0, 133.0, 99.0] 108 144 1.333333 119.0 63.0 1.888889 15552 7497.0 [[[136 167 196]\n [136 167 196]\n [138 169 1...
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
8139 08140.jpg 3.0 44.0 423.0 336.0 78 C:\Users\uprak\capstone\Car Images\Train Image... Chrysler Town and Country Minivan 2012 [3.0, 44.0, 423.0, 336.0] 360 424 1.177778 420.0 292.0 1.438356 152640 122640.0 [[[255 255 255]\n [255 255 255]\n [255 255 2...
8140 08141.jpg 138.0 150.0 706.0 523.0 196 C:\Users\uprak\capstone\Car Images\Train Image... smart fortwo Convertible 2012 [138.0, 150.0, 706.0, 523.0] 600 800 1.333333 568.0 373.0 1.522788 480000 211864.0 [[[ 79 79 53]\n [ 79 79 53]\n [ 79 79 ...
8141 08142.jpg 26.0 246.0 660.0 449.0 163 C:\Users\uprak\capstone\Car Images\Train Image... Mercedes-Benz SL-Class Coupe 2009 [26.0, 246.0, 660.0, 449.0] 469 750 1.599147 634.0 203.0 3.123153 351750 128702.0 [[[108 113 106]\n [111 116 109]\n [111 116 1...
8142 08143.jpg 78.0 526.0 1489.0 908.0 112 C:\Users\uprak\capstone\Car Images\Train Image... Ford GT Coupe 2006 [78.0, 526.0, 1489.0, 908.0] 1067 1600 1.499531 1411.0 382.0 3.693717 1707200 539002.0 [[[ 84 83 78]\n [ 72 72 64]\n [ 59 60 ...
8143 08144.jpg 20.0 240.0 862.0 677.0 17 C:\Users\uprak\capstone\Car Images\Train Image... Audi 100 Sedan 1994 [20.0, 240.0, 862.0, 677.0] 683 883 1.292826 842.0 437.0 1.926773 603089 367954.0 [[[70 70 70]\n [70 70 70]\n [70 70 70]\n .....

8144 rows × 18 columns

In [50]:
df_test['Bbox'] = df_test.apply(lambda x:list([x['Start_x'],x['Start_y'],x['End_x'],x['End_y']]),axis=1)
df_test
Out[50]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Images
0 00001.jpg 30.0 52.0 246.0 147.0 181 C:\Users\uprak\capstone\Car Images\Test Images... Suzuki Aerio Sedan 2007 [30.0, 52.0, 246.0, 147.0] 182 276 [[[254 254 254]\n [254 254 254]\n [254 254 2...
1 00002.jpg 100.0 19.0 576.0 203.0 103 C:\Users\uprak\capstone\Car Images\Test Images... Ferrari 458 Italia Convertible 2012 [100.0, 19.0, 576.0, 203.0] 360 640 [[[126 122 119]\n [124 120 117]\n [121 117 1...
2 00003.jpg 51.0 105.0 968.0 659.0 145 C:\Users\uprak\capstone\Car Images\Test Images... Jeep Patriot SUV 2012 [51.0, 105.0, 968.0, 659.0] 741 1024 [[[161 160 166]\n [159 158 164]\n [157 156 1...
3 00004.jpg 67.0 84.0 581.0 407.0 187 C:\Users\uprak\capstone\Car Images\Test Images... Toyota Camry Sedan 2012 [67.0, 84.0, 581.0, 407.0] 480 640 [[[ 21 21 21]\n [ 24 24 24]\n [ 22 22 ...
4 00005.jpg 140.0 151.0 593.0 339.0 185 C:\Users\uprak\capstone\Car Images\Test Images... Tesla Model S Sedan 2012 [140.0, 151.0, 593.0, 339.0] 373 600 [[[254 254 242]\n [254 254 242]\n [254 254 2...
... ... ... ... ... ... ... ... ... ... ... ... ...
8036 08037.jpg 49.0 57.0 1169.0 669.0 63 C:\Users\uprak\capstone\Car Images\Test Images... Chevrolet Sonic Sedan 2012 [49.0, 57.0, 1169.0, 669.0] 800 1200 [[[ 93 77 77]\n [ 92 76 76]\n [ 91 75 ...
8037 08038.jpg 23.0 18.0 640.0 459.0 16 C:\Users\uprak\capstone\Car Images\Test Images... Audi V8 Sedan 1994 [23.0, 18.0, 640.0, 459.0] 480 640 [[[ 54 51 42]\n [ 50 47 38]\n [ 51 48 ...
8038 08039.jpg 33.0 27.0 602.0 252.0 17 C:\Users\uprak\capstone\Car Images\Test Images... Audi 100 Sedan 1994 [33.0, 27.0, 602.0, 252.0] 303 635 [[[122 83 54]\n [112 75 46]\n [120 83 ...
8039 08040.jpg 33.0 142.0 521.0 376.0 38 C:\Users\uprak\capstone\Car Images\Test Images... BMW Z4 Convertible 2012 [33.0, 142.0, 521.0, 376.0] 427 640 [[[ 74 61 44]\n [ 74 61 44]\n [ 74 61 ...
8040 08041.jpg 77.0 73.0 506.0 380.0 32 C:\Users\uprak\capstone\Car Images\Test Images... BMW X5 SUV 2007 [77.0, 73.0, 506.0, 380.0] 426 640 [[[ 39 39 31]\n [ 71 71 63]\n [ 45 45 ...

8041 rows × 12 columns

In [119]:
df_visualize = df_train.sample(100)
image_paths = list(df_visualize['Image_Path']) 
images = [read_image(p) for p in image_paths]
#images = np.array(df_visualize['Images'])
image_name = list(df_visualize['Car_Name'])
bbox_list = list(df_visualize['Bbox'])
image = visualize(images,bbox_list,image_name)
fig = plt.figure(figsize=(20, 20))
columns = 10
rows = 10

pics = []
for i in range(columns*rows):
    pics.append(fig.add_subplot(rows, columns, i+1,title=image_name[i]))
    plt.imshow(image[i])

plt.show()
In [120]:
df_test_visualize = df_test.sample(100)
image_paths = list(df_test_visualize['Image_Path']) 
images = [read_image(p) for p in image_paths]
#images = np.array(df_visualize['Images'])
image_name = list(df_test_visualize['Car_Name'])
bbox_list = list(df_test_visualize['Bbox'])
image = visualize(images,bbox_list,image_name)
fig = plt.figure(figsize=(20, 20))
columns = 10
rows = 10

pics = []
for i in range(columns*rows):
    pics.append(fig.add_subplot(rows, columns, i+1,title=image_name[i]))
    plt.imshow(image[i])

plt.show()
In [121]:
df_train.describe().transpose()
Out[121]:
count mean std min 25% 50% 75% max
Start_x 8144.0 64.906803 82.198684 1.000000 19.000000 39.000000 79.000000 1.648000e+03
Start_y 8144.0 108.661223 104.551635 1.000000 42.000000 80.000000 138.250000 1.508000e+03
End_x 8144.0 638.208620 410.776734 76.000000 392.000000 569.000000 746.250000 5.205000e+03
End_y 8144.0 416.431606 273.786000 47.000000 248.000000 360.000000 477.000000 3.389000e+03
Image_class 8144.0 98.979371 56.503148 1.000000 50.000000 99.000000 148.000000 1.960000e+02
Orig_img_height 8144.0 482.771979 317.580021 57.000000 290.000000 424.000000 540.000000 3.744000e+03
Orig_img_width 8144.0 699.793099 450.922445 78.000000 429.000000 636.500000 800.000000 5.616000e+03
Img_AR 8144.0 1.473561 0.232987 0.593066 1.333333 1.348485 1.519757 3.887538e+00
bb_width 8144.0 573.301817 375.675050 73.000000 338.000000 504.000000 659.000000 4.156000e+03
bb_height 8144.0 307.770383 213.463238 23.000000 173.000000 260.000000 370.000000 2.648000e+03
bb_AR 8144.0 1.951161 0.488941 0.559322 1.593023 1.867606 2.249385 3.975000e+00
img_area 8144.0 477026.917117 890130.689330 4524.000000 121240.000000 265752.000000 432363.750000 2.102630e+07
bb_area 8144.0 249462.668713 494907.035699 1748.000000 59735.000000 130497.000000 237495.750000 9.874656e+06
In [51]:
df_train['bb_area/img_area'] = df_train['bb_area']/df_train['img_area']
df_train.head(5)
Out[51]:
Image_Name Start_x Start_y End_x End_y Image_class Image_Path Car_Name Bbox Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area Images bb_area/img_area
0 00001.jpg 39.0 116.0 569.0 375.0 14 C:\Users\uprak\capstone\Car Images\Train Image... Audi TTS Coupe 2012 [39.0, 116.0, 569.0, 375.0] 400 600 1.500000 530.0 259.0 2.046332 240000 137270.0 [[[101 119 123]\n [ 98 116 120]\n [ 93 111 1... 0.571958
1 00002.jpg 36.0 116.0 868.0 587.0 3 C:\Users\uprak\capstone\Car Images\Train Image... Acura TL Sedan 2012 [36.0, 116.0, 868.0, 587.0] 675 900 1.333333 832.0 471.0 1.766454 607500 391872.0 [[[164 169 175]\n [166 171 177]\n [169 174 1... 0.645057
2 00003.jpg 85.0 109.0 601.0 381.0 91 C:\Users\uprak\capstone\Car Images\Train Image... Dodge Dakota Club Cab 2007 [85.0, 109.0, 601.0, 381.0] 480 640 1.333333 516.0 272.0 1.897059 307200 140352.0 [[[231 231 231]\n [244 244 244]\n [234 234 2... 0.456875
3 00004.jpg 621.0 393.0 1484.0 1096.0 134 C:\Users\uprak\capstone\Car Images\Train Image... Hyundai Sonata Hybrid Sedan 2012 [621.0, 393.0, 1484.0, 1096.0] 1386 2100 1.515152 863.0 703.0 1.227596 2910600 606689.0 [[[255 255 255]\n [255 255 255]\n [255 255 2... 0.208441
4 00005.jpg 14.0 36.0 133.0 99.0 106 C:\Users\uprak\capstone\Car Images\Train Image... Ford F-450 Super Duty Crew Cab 2012 [14.0, 36.0, 133.0, 99.0] 108 144 1.333333 119.0 63.0 1.888889 15552 7497.0 [[[136 167 196]\n [136 167 196]\n [138 169 1... 0.482060
In [52]:
cols = list(df_train.columns)
cols
Out[52]:
['Image_Name',
 'Start_x',
 'Start_y',
 'End_x',
 'End_y',
 'Image_class',
 'Image_Path',
 'Car_Name',
 'Bbox',
 'Orig_img_height',
 'Orig_img_width',
 'Img_AR',
 'bb_width',
 'bb_height',
 'bb_AR',
 'img_area',
 'bb_area',
 'Images',
 'bb_area/img_area']
In [53]:
df_eda = df_train[['Image_Name','Image_class','Bbox','Orig_img_width','Orig_img_height','Img_AR','bb_width','bb_height','bb_AR',
                  'img_area','bb_area','bb_area/img_area']]
In [54]:
df_eda.sample(10)
Out[54]:
Image_Name Image_class Bbox Orig_img_width Orig_img_height Img_AR bb_width bb_height bb_AR img_area bb_area bb_area/img_area
2672 02673.jpg 59 [72.0, 61.0, 541.0, 291.0] 550 323 1.702786 469.0 230.0 2.039130 177650 107870.0 0.607205
3318 03319.jpg 125 [19.0, 35.0, 242.0, 174.0] 259 194 1.335052 223.0 139.0 1.604317 50246 30997.0 0.616905
3178 03179.jpg 149 [196.0, 113.0, 1563.0, 1118.0] 1920 1280 1.500000 1367.0 1005.0 1.360199 2457600 1373835.0 0.559015
5680 05681.jpg 77 [55.0, 84.0, 393.0, 278.0] 451 338 1.334320 338.0 194.0 1.742268 152438 65572.0 0.430155
1108 01109.jpg 167 [578.0, 169.0, 1427.0, 925.0] 1921 1080 1.778704 849.0 756.0 1.123016 2074680 641844.0 0.309370
5775 05776.jpg 145 [11.0, 87.0, 944.0, 736.0] 1024 768 1.333333 933.0 649.0 1.437596 786432 605517.0 0.769955
2258 02259.jpg 44 [14.0, 19.0, 450.0, 274.0] 454 280 1.621429 436.0 255.0 1.709804 127120 111180.0 0.874607
1356 01357.jpg 44 [52.0, 122.0, 750.0, 429.0] 800 533 1.500938 698.0 307.0 2.273616 426400 214286.0 0.502547
5184 05185.jpg 112 [80.0, 231.0, 861.0, 679.0] 1024 768 1.333333 781.0 448.0 1.743304 786432 349888.0 0.444906
2024 02025.jpg 22 [10.0, 42.0, 310.0, 180.0] 320 200 1.600000 300.0 138.0 2.173913 64000 41400.0 0.646875
In [55]:
sns.histplot(data=df_eda,x='Img_AR',binwidth=0.1)
Out[55]:
<AxesSubplot:xlabel='Img_AR', ylabel='Count'>
In [56]:
np.mean(np.array(df_eda['Img_AR']))
Out[56]:
1.4735606114201925
In [57]:
sns.histplot(data=df_eda,x='bb_AR',binwidth=0.2)
Out[57]:
<AxesSubplot:xlabel='bb_AR', ylabel='Count'>
In [58]:
np.mean(np.array(df_eda['bb_AR']))
Out[58]:
1.9511606317887489
In [59]:
df_group.sample(5)
Out[59]:
Start_x Start_y End_x End_y Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area
Image_class
134 75.823529 122.676471 573.294118 370.352941 444.529412 665.764706 1.523828 497.470588 247.676471 2.065191 399625.382353 154786.264706
119 64.161765 85.779412 610.647059 403.235294 459.088235 664.720588 1.480158 546.485294 317.455882 1.784687 420236.470588 233624.941176
76 39.000000 65.045455 439.840909 298.681818 335.909091 483.022727 1.473433 400.840909 233.636364 1.747487 237099.272727 135174.136364
98 70.152174 154.326087 894.847826 581.782609 695.695652 974.000000 1.422008 824.695652 427.456522 2.017653 974127.000000 498596.217391
39 98.972973 118.351351 526.540541 334.729730 391.540541 599.351351 1.542289 427.567568 216.378378 2.101122 329751.486486 128508.108108
In [60]:
df_group.reset_index(inplace=True)
df_group
Out[60]:
Image_class Start_x Start_y End_x End_y Orig_img_height Orig_img_width Img_AR bb_width bb_height bb_AR img_area bb_area
0 1 33.355556 67.333333 449.688889 306.000000 338.777778 491.133333 1.488229 416.333333 238.666667 1.733447 339422.000000 185725.533333
1 2 69.718750 113.250000 633.250000 380.281250 444.343750 694.250000 1.626136 563.531250 267.031250 2.193562 421385.937500 192106.718750
2 3 63.953488 99.046512 657.558140 419.860465 482.976744 714.046512 1.490232 593.604651 320.813953 1.999234 510487.860465 294358.186047
3 4 50.380952 131.261905 673.285714 435.357143 505.428571 720.023810 1.427891 622.904762 304.095238 2.079940 416349.880952 211456.714286
4 5 47.195122 95.853659 486.926829 313.414634 370.536585 529.878049 1.467877 439.731707 217.560976 2.016969 346084.951220 169860.097561
... ... ... ... ... ... ... ... ... ... ... ... ... ...
191 192 116.395349 163.162791 912.906977 600.930233 699.348837 1021.767442 1.480585 796.511628 437.767442 1.924556 825807.209302 433913.860465
192 193 83.833333 113.166667 693.047619 454.404762 517.261905 778.547619 1.541277 609.214286 341.238095 1.873040 487950.666667 241788.619048
193 194 84.391304 123.260870 787.304348 485.195652 566.913043 837.391304 1.539087 702.913043 361.934783 2.088240 537252.239130 285973.065217
194 195 54.674419 74.581395 550.883721 367.627907 409.813953 595.023256 1.478335 496.209302 293.046512 1.713090 288334.674419 171179.627907
195 196 62.750000 70.550000 553.850000 391.575000 426.250000 623.525000 1.462446 491.100000 321.025000 1.526207 373605.325000 218841.350000

196 rows × 13 columns

In [61]:
df_group['Car_Name'] = df_group['Image_class'].apply(lambda x:car_dict[int(x)])
In [62]:
sns.barplot(data=df_group,x='Image_class',y='bb_area')
Out[62]:
<AxesSubplot:xlabel='Image_class', ylabel='bb_area'>
In [63]:
np.mean(np.array(df_eda['bb_area']))
Out[63]:
249462.66871316306
In [153]:
target = os.path.join(car_img_path,'train_img_resize')
image_paths = list(df_train['Image_Path']) 
image_name = list(df_train['Image_Name'])
i=0
for p in image_paths:
    to_file = os.path.join(target,image_name[i])
    im = read_image(p)
    image = np.array(tf.image.resize_with_pad(im,224,224,method=tf.image.ResizeMethod.BILINEAR,antialias=False))
    #img = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
    image = array_to_img(image)
    image.save(to_file)
    #cv.imwrite(target,image,[cv.IMWRITE_JPEG_QUALITY, 50])
#    if i == 0:
#        imgs = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
#    else:
#        img = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
#        imgs = np.append(imgs,img,axis=0)  
    i = i + 1
In [154]:
target = os.path.join(car_img_path,'test_img_resize')
image_paths = list(df_test['Image_Path']) 
image_name = list(df_test['Image_Name'])
i=0
for p in image_paths:
    to_file = os.path.join(target,image_name[i])
    im = read_image(p)
    image = np.array(tf.image.resize_with_pad(im,224,224,method=tf.image.ResizeMethod.BILINEAR,antialias=False))
    #img = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
    image = array_to_img(image)
    image.save(to_file)
    #cv.imwrite(target,image,[cv.IMWRITE_JPEG_QUALITY, 50])
#    if i == 0:
#        imgs = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
#    else:
#        img = np.expand_dims(np.array(image,dtype=float)/255,axis=0)
#        imgs = np.append(imgs,img,axis=0)  
    i = i + 1
In [64]:
f_train_name = {}
file_path = os.path.join(car_img_path,'train_img_resize')
for filename in os.listdir(file_path):
    f_train_name[filename] = os.path.join(file_path,filename)
In [65]:
f_test_name = {}
file_path = os.path.join(car_img_path,'test_img_resize')
for filename in os.listdir(file_path):
    f_test_name[filename] = os.path.join(file_path,filename)
In [66]:
df_train['Resize_Image_Path'] = df_train['Image_Name'].apply(lambda x:f_train_name[x])
df_test['Resize_Image_Path'] = df_test['Image_Name'].apply(lambda x:f_test_name[x])
In [67]:
df_test['Resize_Image_Path']
Out[67]:
0       C:\Users\uprak\capstone\Car Images\test_img_re...
1       C:\Users\uprak\capstone\Car Images\test_img_re...
2       C:\Users\uprak\capstone\Car Images\test_img_re...
3       C:\Users\uprak\capstone\Car Images\test_img_re...
4       C:\Users\uprak\capstone\Car Images\test_img_re...
                              ...                        
8036    C:\Users\uprak\capstone\Car Images\test_img_re...
8037    C:\Users\uprak\capstone\Car Images\test_img_re...
8038    C:\Users\uprak\capstone\Car Images\test_img_re...
8039    C:\Users\uprak\capstone\Car Images\test_img_re...
8040    C:\Users\uprak\capstone\Car Images\test_img_re...
Name: Resize_Image_Path, Length: 8041, dtype: object
In [68]:
ytrain = list(df_train['Car_Name'])
ytest = list(df_test['Car_Name'])
In [69]:
train_datagen = ImageDataGenerator(rescale=1./255,validation_split=0.25)
test_datagen  = ImageDataGenerator(rescale=1./255)
train_generator = train_datagen.flow_from_dataframe(df_train,directory=None, x_col='Resize_Image_Path', y_col='Car_Name',subset="training",weight_col=None,target_size=(224,224),color_mode='rgb',class_mode='categorical',batch_size=8,shuffle=True)
valid_generator = train_datagen.flow_from_dataframe(df_train,directory=None, x_col='Resize_Image_Path', y_col='Car_Name',subset="validation",weight_col=None,target_size=(224,224),color_mode='rgb',class_mode='categorical',batch_size=8,shuffle=True)
test_generator = test_datagen.flow_from_dataframe(df_test,directory=None, x_col='Resize_Image_Path', y_col=None,weight_col=None,target_size=(224,224),color_mode='rgb',class_mode=None,batch_size=8,shuffle=False)

Found 6108 validated image filenames belonging to 196 classes.
Found 2036 validated image filenames belonging to 196 classes.
Found 8041 validated image filenames.
In [70]:
class_label_map = train_generator.class_indices
class_label_map
Out[70]:
{'AM General Hummer SUV 2000': 0,
 'Acura Integra Type R 2001': 1,
 'Acura RL Sedan 2012': 2,
 'Acura TL Sedan 2012': 3,
 'Acura TL Type-S 2008': 4,
 'Acura TSX Sedan 2012': 5,
 'Acura ZDX Hatchback 2012': 6,
 'Aston Martin V8 Vantage Convertible 2012': 7,
 'Aston Martin V8 Vantage Coupe 2012': 8,
 'Aston Martin Virage Convertible 2012': 9,
 'Aston Martin Virage Coupe 2012': 10,
 'Audi 100 Sedan 1994': 11,
 'Audi 100 Wagon 1994': 12,
 'Audi A5 Coupe 2012': 13,
 'Audi R8 Coupe 2012': 14,
 'Audi RS 4 Convertible 2008': 15,
 'Audi S4 Sedan 2007': 16,
 'Audi S4 Sedan 2012': 17,
 'Audi S5 Convertible 2012': 18,
 'Audi S5 Coupe 2012': 19,
 'Audi S6 Sedan 2011': 20,
 'Audi TT Hatchback 2011': 21,
 'Audi TT RS Coupe 2012': 22,
 'Audi TTS Coupe 2012': 23,
 'Audi V8 Sedan 1994': 24,
 'BMW 1 Series Convertible 2012': 25,
 'BMW 1 Series Coupe 2012': 26,
 'BMW 3 Series Sedan 2012': 27,
 'BMW 3 Series Wagon 2012': 28,
 'BMW 6 Series Convertible 2007': 29,
 'BMW ActiveHybrid 5 Sedan 2012': 30,
 'BMW M3 Coupe 2012': 31,
 'BMW M5 Sedan 2010': 32,
 'BMW M6 Convertible 2010': 33,
 'BMW X3 SUV 2012': 34,
 'BMW X5 SUV 2007': 35,
 'BMW X6 SUV 2012': 36,
 'BMW Z4 Convertible 2012': 37,
 'Bentley Arnage Sedan 2009': 38,
 'Bentley Continental Flying Spur Sedan 2007': 39,
 'Bentley Continental GT Coupe 2007': 40,
 'Bentley Continental GT Coupe 2012': 41,
 'Bentley Continental Supersports Conv. Convertible 2012': 42,
 'Bentley Mulsanne Sedan 2011': 43,
 'Bugatti Veyron 16.4 Convertible 2009': 44,
 'Bugatti Veyron 16.4 Coupe 2009': 45,
 'Buick Enclave SUV 2012': 46,
 'Buick Rainier SUV 2007': 47,
 'Buick Regal GS 2012': 48,
 'Buick Verano Sedan 2012': 49,
 'Cadillac CTS-V Sedan 2012': 50,
 'Cadillac Escalade EXT Crew Cab 2007': 51,
 'Cadillac SRX SUV 2012': 52,
 'Chevrolet Avalanche Crew Cab 2012': 53,
 'Chevrolet Camaro Convertible 2012': 54,
 'Chevrolet Cobalt SS 2010': 55,
 'Chevrolet Corvette Convertible 2012': 56,
 'Chevrolet Corvette Ron Fellows Edition Z06 2007': 57,
 'Chevrolet Corvette ZR1 2012': 58,
 'Chevrolet Express Cargo Van 2007': 59,
 'Chevrolet Express Van 2007': 60,
 'Chevrolet HHR SS 2010': 61,
 'Chevrolet Impala Sedan 2007': 62,
 'Chevrolet Malibu Hybrid Sedan 2010': 63,
 'Chevrolet Malibu Sedan 2007': 64,
 'Chevrolet Monte Carlo Coupe 2007': 65,
 'Chevrolet Silverado 1500 Classic Extended Cab 2007': 66,
 'Chevrolet Silverado 1500 Extended Cab 2012': 67,
 'Chevrolet Silverado 1500 Hybrid Crew Cab 2012': 68,
 'Chevrolet Silverado 1500 Regular Cab 2012': 69,
 'Chevrolet Silverado 2500HD Regular Cab 2012': 70,
 'Chevrolet Sonic Sedan 2012': 71,
 'Chevrolet Tahoe Hybrid SUV 2012': 72,
 'Chevrolet TrailBlazer SS 2009': 73,
 'Chevrolet Traverse SUV 2012': 74,
 'Chrysler 300 SRT-8 2010': 75,
 'Chrysler Aspen SUV 2009': 76,
 'Chrysler Crossfire Convertible 2008': 77,
 'Chrysler PT Cruiser Convertible 2008': 78,
 'Chrysler Sebring Convertible 2010': 79,
 'Chrysler Town and Country Minivan 2012': 80,
 'Daewoo Nubira Wagon 2002': 81,
 'Dodge Caliber Wagon 2007': 82,
 'Dodge Caliber Wagon 2012': 83,
 'Dodge Caravan Minivan 1997': 84,
 'Dodge Challenger SRT8 2011': 85,
 'Dodge Charger SRT-8 2009': 86,
 'Dodge Charger Sedan 2012': 87,
 'Dodge Dakota Club Cab 2007': 88,
 'Dodge Dakota Crew Cab 2010': 89,
 'Dodge Durango SUV 2007': 90,
 'Dodge Durango SUV 2012': 91,
 'Dodge Journey SUV 2012': 92,
 'Dodge Magnum Wagon 2008': 93,
 'Dodge Ram Pickup 3500 Crew Cab 2010': 94,
 'Dodge Ram Pickup 3500 Quad Cab 2009': 95,
 'Dodge Sprinter Cargo Van 2009': 96,
 'Eagle Talon Hatchback 1998': 97,
 'FIAT 500 Abarth 2012': 98,
 'FIAT 500 Convertible 2012': 99,
 'Ferrari 458 Italia Convertible 2012': 100,
 'Ferrari 458 Italia Coupe 2012': 101,
 'Ferrari California Convertible 2012': 102,
 'Ferrari FF Coupe 2012': 103,
 'Fisker Karma Sedan 2012': 104,
 'Ford E-Series Wagon Van 2012': 105,
 'Ford Edge SUV 2012': 106,
 'Ford Expedition EL SUV 2009': 107,
 'Ford F-150 Regular Cab 2007': 108,
 'Ford F-150 Regular Cab 2012': 109,
 'Ford F-450 Super Duty Crew Cab 2012': 110,
 'Ford Fiesta Sedan 2012': 111,
 'Ford Focus Sedan 2007': 112,
 'Ford Freestar Minivan 2007': 113,
 'Ford GT Coupe 2006': 114,
 'Ford Mustang Convertible 2007': 115,
 'Ford Ranger SuperCab 2011': 116,
 'GMC Acadia SUV 2012': 117,
 'GMC Canyon Extended Cab 2012': 118,
 'GMC Savana Van 2012': 119,
 'GMC Terrain SUV 2012': 120,
 'GMC Yukon Hybrid SUV 2012': 121,
 'Geo Metro Convertible 1993': 122,
 'HUMMER H2 SUT Crew Cab 2009': 123,
 'HUMMER H3T Crew Cab 2010': 124,
 'Honda Accord Coupe 2012': 125,
 'Honda Accord Sedan 2012': 126,
 'Honda Odyssey Minivan 2007': 127,
 'Honda Odyssey Minivan 2012': 128,
 'Hyundai Accent Sedan 2012': 129,
 'Hyundai Azera Sedan 2012': 130,
 'Hyundai Elantra Sedan 2007': 131,
 'Hyundai Elantra Touring Hatchback 2012': 132,
 'Hyundai Genesis Sedan 2012': 133,
 'Hyundai Santa Fe SUV 2012': 134,
 'Hyundai Sonata Hybrid Sedan 2012': 135,
 'Hyundai Sonata Sedan 2012': 136,
 'Hyundai Tucson SUV 2012': 137,
 'Hyundai Veloster Hatchback 2012': 138,
 'Hyundai Veracruz SUV 2012': 139,
 'Infiniti G Coupe IPL 2012': 140,
 'Infiniti QX56 SUV 2011': 141,
 'Isuzu Ascender SUV 2008': 142,
 'Jaguar XK XKR 2012': 143,
 'Jeep Compass SUV 2012': 144,
 'Jeep Grand Cherokee SUV 2012': 145,
 'Jeep Liberty SUV 2012': 146,
 'Jeep Patriot SUV 2012': 147,
 'Jeep Wrangler SUV 2012': 148,
 'Lamborghini Aventador Coupe 2012': 149,
 'Lamborghini Diablo Coupe 2001': 150,
 'Lamborghini Gallardo LP 570-4 Superleggera 2012': 151,
 'Lamborghini Reventon Coupe 2008': 152,
 'Land Rover LR2 SUV 2012': 153,
 'Land Rover Range Rover SUV 2012': 154,
 'Lincoln Town Car Sedan 2011': 155,
 'MINI Cooper Roadster Convertible 2012': 156,
 'Maybach Landaulet Convertible 2012': 157,
 'Mazda Tribute SUV 2011': 158,
 'McLaren MP4-12C Coupe 2012': 159,
 'Mercedes-Benz 300-Class Convertible 1993': 160,
 'Mercedes-Benz C-Class Sedan 2012': 161,
 'Mercedes-Benz E-Class Sedan 2012': 162,
 'Mercedes-Benz S-Class Sedan 2012': 163,
 'Mercedes-Benz SL-Class Coupe 2009': 164,
 'Mercedes-Benz Sprinter Van 2012': 165,
 'Mitsubishi Lancer Sedan 2012': 166,
 'Nissan 240SX Coupe 1998': 167,
 'Nissan Juke Hatchback 2012': 168,
 'Nissan Leaf Hatchback 2012': 169,
 'Nissan NV Passenger Van 2012': 170,
 'Plymouth Neon Coupe 1999': 171,
 'Porsche Panamera Sedan 2012': 172,
 'Ram C/V Cargo Van Minivan 2012': 173,
 'Rolls-Royce Ghost Sedan 2012': 174,
 'Rolls-Royce Phantom Drophead Coupe Convertible 2012': 175,
 'Rolls-Royce Phantom Sedan 2012': 176,
 'Scion xD Hatchback 2012': 177,
 'Spyker C8 Convertible 2009': 178,
 'Spyker C8 Coupe 2009': 179,
 'Suzuki Aerio Sedan 2007': 180,
 'Suzuki Kizashi Sedan 2012': 181,
 'Suzuki SX4 Hatchback 2012': 182,
 'Suzuki SX4 Sedan 2012': 183,
 'Tesla Model S Sedan 2012': 184,
 'Toyota 4Runner SUV 2012': 185,
 'Toyota Camry Sedan 2012': 186,
 'Toyota Corolla Sedan 2012': 187,
 'Toyota Sequoia SUV 2012': 188,
 'Volkswagen Beetle Hatchback 2012': 189,
 'Volkswagen Golf Hatchback 1991': 190,
 'Volkswagen Golf Hatchback 2012': 191,
 'Volvo 240 Sedan 1993': 192,
 'Volvo C30 Hatchback 2012': 193,
 'Volvo XC90 SUV 2007': 194,
 'smart fortwo Convertible 2012': 195}
In [71]:
model = Sequential()
model.add(Conv2D(32, (3, 3), padding='same',
                 input_shape=(224,224,3)))
model.add(Activation('relu'))
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(64, (3, 3), padding='same'))
model.add(Activation('relu'))
model.add(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(512))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(196, activation='softmax'))
model.compile(optimizers.RMSprop(learning_rate=0.0001, decay=1e-6),loss="categorical_crossentropy",metrics=["accuracy"])
In [72]:
STEP_SIZE_TRAIN=train_generator.n//train_generator.batch_size
STEP_SIZE_VALID=valid_generator.n//valid_generator.batch_size
STEP_SIZE_TEST=test_generator.n//test_generator.batch_size
model.fit_generator(generator=train_generator,
                    steps_per_epoch=STEP_SIZE_TRAIN,
                    validation_data=valid_generator,
                    validation_steps=STEP_SIZE_VALID,
                    epochs=20)

Epoch 1/20

---------------------------------------------------------------------------
ResourceExhaustedError                    Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_20240/3875776026.py in <module>
      2 STEP_SIZE_VALID=valid_generator.n//valid_generator.batch_size
      3 STEP_SIZE_TEST=test_generator.n//test_generator.batch_size
----> 4 model.fit_generator(generator=train_generator,
      5                     steps_per_epoch=STEP_SIZE_TRAIN,
      6                     validation_data=valid_generator,

~\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py in fit_generator(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, validation_freq, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch)
   2014         'Please use `Model.fit`, which supports generators.',
   2015         stacklevel=2)
-> 2016     return self.fit(
   2017         generator,
   2018         steps_per_epoch=steps_per_epoch,

~\anaconda3\envs\tflow\lib\site-packages\keras\utils\traceback_utils.py in error_handler(*args, **kwargs)
     65     except Exception as e:  # pylint: disable=broad-except
     66       filtered_tb = _process_traceback_frames(e.__traceback__)
---> 67       raise e.with_traceback(filtered_tb) from None
     68     finally:
     69       del filtered_tb

~\anaconda3\envs\tflow\lib\site-packages\tensorflow\python\eager\execute.py in quick_execute(op_name, num_outputs, inputs, attrs, ctx, name)
     56   try:
     57     ctx.ensure_initialized()
---> 58     tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,
     59                                         inputs, attrs, num_outputs)
     60   except core._NotOkStatusException as e:

ResourceExhaustedError:  failed to allocate memory
	 [[node RMSprop/RMSprop/update_8/mul_2
 (defined at C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\rmsprop.py:211)
]]
Hint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info. This isn't available when running in Eager mode.
 [Op:__inference_train_function_1300]

Errors may have originated from an input operation.
Input Source operations connected to node RMSprop/RMSprop/update_8/mul_2:
In[0] RMSprop/Identity (defined at C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\optimizer_v2.py:955)	
In[1] RMSprop/IdentityN (defined at C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\utils.py:35)

Operation defined at: (most recent call last)
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\runpy.py", line 197, in _run_module_as_main
>>>     return _run_code(code, main_globals, None,
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\runpy.py", line 87, in _run_code
>>>     exec(code, run_globals)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel_launcher.py", line 16, in <module>
>>>     app.launch_new_instance()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\traitlets\config\application.py", line 845, in launch_instance
>>>     app.start()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\kernelapp.py", line 677, in start
>>>     self.io_loop.start()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\tornado\platform\asyncio.py", line 199, in start
>>>     self.asyncio_loop.run_forever()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\asyncio\base_events.py", line 596, in run_forever
>>>     self._run_once()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\asyncio\base_events.py", line 1890, in _run_once
>>>     handle._run()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\asyncio\events.py", line 80, in _run
>>>     self._context.run(self._callback, *self._args)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\kernelbase.py", line 457, in dispatch_queue
>>>     await self.process_one()
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\kernelbase.py", line 446, in process_one
>>>     await dispatch(*args)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\kernelbase.py", line 353, in dispatch_shell
>>>     await result
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\kernelbase.py", line 648, in execute_request
>>>     reply_content = await reply_content
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\ipkernel.py", line 353, in do_execute
>>>     res = shell.run_cell(code, store_history=store_history, silent=silent)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\ipykernel\zmqshell.py", line 533, in run_cell
>>>     return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\interactiveshell.py", line 2901, in run_cell
>>>     result = self._run_cell(
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\interactiveshell.py", line 2947, in _run_cell
>>>     return runner(coro)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\async_helpers.py", line 68, in _pseudo_sync_runner
>>>     coro.send(None)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\interactiveshell.py", line 3172, in run_cell_async
>>>     has_raised = await self.run_ast_nodes(code_ast.body, cell_name,
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\interactiveshell.py", line 3364, in run_ast_nodes
>>>     if (await self.run_code(code, result,  async_=asy)):
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\IPython\core\interactiveshell.py", line 3444, in run_code
>>>     exec(code_obj, self.user_global_ns, self.user_ns)
>>> 
>>>   File "C:\Users\uprak\AppData\Local\Temp/ipykernel_20240/3875776026.py", line 4, in <module>
>>>     model.fit_generator(generator=train_generator,
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 2016, in fit_generator
>>>     return self.fit(
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\utils\traceback_utils.py", line 64, in error_handler
>>>     return fn(*args, **kwargs)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 1216, in fit
>>>     tmp_logs = self.train_function(iterator)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 878, in train_function
>>>     return step_function(self, iterator)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 867, in step_function
>>>     outputs = model.distribute_strategy.run(run_step, args=(data,))
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 860, in run_step
>>>     outputs = model.train_step(data)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\engine\training.py", line 816, in train_step
>>>     self.optimizer.minimize(loss, self.trainable_variables, tape=tape)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\optimizer_v2.py", line 532, in minimize
>>>     return self.apply_gradients(grads_and_vars, name=name)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\optimizer_v2.py", line 672, in apply_gradients
>>>     return self._distributed_apply(strategy, grads_and_vars, name,
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\optimizer_v2.py", line 721, in _distributed_apply
>>>     update_op = distribution.extended.update(
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\optimizer_v2.py", line 704, in apply_grad_to_update_var
>>>     update_op = self._resource_apply_dense(grad, var, **apply_kwargs)
>>> 
>>>   File "C:\Users\uprak\anaconda3\envs\tflow\lib\site-packages\keras\optimizer_v2\rmsprop.py", line 211, in _resource_apply_dense
>>>     var_t = var - coefficients["lr_t"] * grad / (
>>>
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]: